AI Is Making Workers More Productive. So Where Is the Enterprise ROI?
AI is working—at least, that is what employees are reporting. In McKinsey’s 2026 State of AI survey, 80% of respondents said AI had improved their individual productivity, with roughly half reporting improved decision-making and skill development.
That sounds like a major business success, but then the picture changes: only 37% of respondents said AI had contributed positively to their organization’s EBIT—a figure essentially unchanged from the year before. Just 6% qualified as "AI high performers" (attributing at least 5% of EBIT to AI while reporting significant value).
So we have a central contradiction:
AI appears to be making people more productive much faster than it is making companies more profitable.
Why? That may be one of the most important questions in enterprise AI right now.
Individual Productivity Is Not the Same as Enterprise Value
Across organizational roles, roughly four out of five respondents reported productivity improvements from AI, along with benefits in decision-making, skill development, creativity, and clarity of thought.
Those gains matter, but an employee completing a task 20% faster does not automatically create a measurable financial return for the enterprise. The critical question is what happens to the saved time:
- Does the organization increase output, reduce costs, or serve more customers?
- Does it launch products faster, improve quality, or reduce headcount?
- Does it redeploy people to higher-value work?
- Or does the time simply disappear into the operating system?
Productivity is a capability gain, whereas ROI requires converting that capability into a measurable economic outcome.
The Value Is Real—but Fragmented
McKinsey’s survey shows clear financial benefits inside individual business functions:
Most Common Cost Reductions: Supply chain management, service operations, and manufacturing.
Most Common Revenue Increases: Marketing and sales, product/service development, and software engineering.
The problem is not that AI produces no value; it is producing value in many localized areas. The harder problem is that this value remains:
Localized → Fragmented → Inconsistently Measured → Difficult to Attribute → Difficult to Aggregate
A company can have ten successful use cases and still struggle to determine whether its AI portfolio creates enough enterprise-level value to justify total spending.
Scaling Is Getting Ahead of Value Realization
Nearly nine in ten respondents in McKinsey’s 2026 survey said their organizations regularly use AI in at least one business function. Among organizations using AI, the share reaching the scaling phase rose from 38% in 2025 to 44% in 2026.
For large enterprises (over $1 billion in annual revenue), the share scaling AI agents increased from 27% to 40% in just one year. Organizations are not waiting for the ROI framework to be fully solved before scaling up.
AI deployment can expand faster than the organization's ability to determine whether the deployment is economically working.
Why Productivity May Not Reach the Bottom Line
Some benefits take time to appear financially or show up through customer satisfaction and competitive positioning rather than immediate EBIT. Organizations may also be absorbing heavy implementation and operating costs while benefits accumulate.
However, McKinsey’s high performers provide a vital clue: nearly three-quarters of AI high performers report fundamentally redesigning workflows because of AI, compared with roughly one-quarter of other respondents. High performers are also much more likely to measure AI impact, involve senior leadership, use human-in-the-loop designs, actively manage costs, and mitigate risks.
The missing link between individual productivity and enterprise ROI may be the operating model around the technology.
Faster Work Does Not Automatically Mean a Better Workflow
Suppose an employee previously spent ten hours preparing a report, and AI reduces that work to five hours—a 50% productivity improvement. The financial result depends entirely on what happens next:
Unchanged Workflow: The company simply gets the same report completed faster.
Redesigned Process: The saved hours are structured to produce deeper analysis, eliminate unnecessary steps, consolidate software tools, serve more customers, or eliminate external spending.
The productivity gain remains identical; the management of that gain determines whether it hits the bottom line.
Measuring AI Impact Changes the Decision
High performers are substantially more likely to have defined processes for quantifying the impact of AI initiatives. Yet many organizations still measure AI through activity metrics: users, prompts, licenses, tokens, pilots, models, and agents launched.
An effective AI management system must track outcome-oriented questions:
- What outcome was supposed to improve, and what was the baseline?
- What changed, and how much can reasonably be attributed to AI?
- What did the system cost, and is performance improving over time?
- What decision should leadership make because of that evidence?
From Measurement to Management
AI ROI must evolve from a static reporting exercise into an active decision architecture:
AI Initiative → Expected Value → Realized Value → Cost → Risk → Confidence → Trend → Decision
That decision loop dictates explicit actions: Scale | Continue | Govern | Redesign | Route to Cheaper Workflow | Pause | Stop. A dashboard tells leadership what happened; a decision system helps leadership determine what to do next.
The Decision Systems Approach
To bridge this gap, Decision Systems Orchestrators connect isolated AI outputs to broader business infrastructure:
- Business metrics & cost controls
- Deterministic rules & threshold checks
- Risk & confidence evaluation
- Ownership & escalation logic
- Explicit action routing
The objective is not merely to make the model smarter, but to make the business decision surrounding the model accountable.
High Performers Are Showing What Comes Next
McKinsey describes the primary differentiator for high performers as the coherence of their approach. Organizations generating strong AI value do not simply accumulate more tools; they build operating models capable of evaluating:
Where does AI belong? What should change around it? How do we measure the result? When should we scale, redesign, or stop?
The Next AI Bottleneck Is Organizational
AI capabilities continue to advance rapidly, adoption is rising, and employees are gaining productivity, but the limiting factor is increasingly an organization's ability to absorb change, redesign workflows, and build the operating capability required to scale.
The core executive question is no longer “Are our employees becoming more productive with AI?”
The critical question is:
Do we have a system for turning that productivity into measurable business value—and deciding what to do next?